The goal of this three year proposal is to develop E. coli as an efficient heterologous expression system for the DNA bisintercalator triostin. A non-ribosomally synthesized peptide, triostin is a known antibiotic and anticancer agent. We will test the hypothesis that by metabolically engineering the E. coli host, modifying the plasmid vector, and altering the triostin gene, triostin and its analogs could be biosynthesized. There are two specific aims for this research, (1) Metabolically engineer E. coli to produce triostin. (2) Modify the triostin NRPS to produce analogs with altered sequence specificity and selectivity. Preliminary results. We have identified a four module non-ribosomal peptide synthetase that contains adenylation domains specific for Serine, Alanine, Cysteine, Valine amino acids (the four amino acids found in triostin.) The significance of this research is that it will (a) improve the mechanistic understanding of the assembly of C2-symmetric non-ribosomal peptides, (b) provide a general host for the expression of diverse non-ribosomal peptide synthetase genes.